Approximations to the truth: comparing survey and microsimulation approaches to measuring income for social indicators

Abstract

In this paper, we evaluate income distributions in four European countries (Austria, Italy, Spain and Hungary) using two complementary approaches: a standard approach based on reported incomes in survey data, and a microsimulation approach, where taxes and benefits are simulated. These two approaches may be expected to generate slightly different results, particularly in respect of individuals on lower incomes, because benefit receipts tend to be under-reported in survey data, and over-estimated in microsimulation procedures. However, we find that the two approaches do in fact produce reasonably consistent results, in terms of both inequality measures and poverty rates. To the extent that the results differ, we explore the reasons why these differences arise, and suggest directions for future research, in which each approach may inform improvements in the other.